Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2503.22269 · Physical Review A

Hardware-efficient quantum annealing with error mitigation via classical shadow

Takaharu Yoshida, Yuta Shingu, Chihaya Shimada, Tetsuro Nikuni, Hideaki Hakoshima, Yuichiro Matsuzaki

Published Apr 21, 2026• 1 citation
arXivHTMLPDF

Abstract

Quantum annealing (QA) is an efficient method for finding the ground-state energy of the problem Hamiltonian. However, in practical implementation, the system suffers from decoherence. On the other hand, recently, ``Localized virtual purification" (LVP) was proposed to suppress decoherence in the context of noisy intermediate-scale quantum (NISQ) devices. Suppose observables have spatially local support in the lattice. In that case, the requirement for LVP is to calculate the expectation value with a reduced density matrix on a portion of the total system. In this work, we propose a method to mitigate decoherence errors in QA using LVP. The key idea is to use the so-called classical shadow method to construct the reduced density matrix. Thanks to the CS, unlike the previous schemes to mitigate decoherence error for QA, we do not need either two-qubit gates or mid-circuit measurements, which means that our method is hardware-efficient.

DOI
10.1103/stb4-wms8

Related papers

Quantum annealing with error mitigation

Yuta Shingu, Tetsuro Nikuni, Shiro Kawabata +1 more

Quantum annealing (QA) is one of the methods to prepare a ground state of a problem Hamiltonian. In the absence of noise, QA can accurately estimate the ground-state energy if the adiabatic condition is satisfied. However, in practice, systems are known to suffer from decoherence. Meanwhile, considerable research effort has been devoted to noisy intermediate-scale quantum (NISQ) computation. For practical NISQ computation, many error-mitigation (EM) methods have been proposed to mitigate the effects of noise. This paper proposes a QA strategy combined with an EM method, namely, dual-state purification, to suppress the effects of decoherence. Our protocol consists of four parts: the conventional dynamics, single-qubit projective measurements, the Hamiltonian dynamics corresponding to an inverse map of the first dynamics, and postprocessing of the measurement results. Importantly, our protocol works without two-qubit gates that require pulse operations; hence, it is suitable for devices designed for practical QA. In addition, we present numerical calculations to show that our protocol is more accurate than the conventional QA in estimating the ground-state energy under decoherence.

arXiv:2210.08862•Physical Review A•Apr 2, 2024•9 citations

Enhancing quantum annealing accuracy through replication-based error mitigation*

Hristo N. Djidjev

Abstract Quantum annealers like those manufactured by D-Wave Systems are designed to find high quality solutions to optimization problems that are typically hard for classical computers. They utilize quantum effects like tunneling to evolve toward low-energy states representing solutions to optimization problems. However, their analog nature and limited control functionalities present challenges to correcting or mitigating hardware errors. As quantum computing advances towards applications, effective error suppression is an important research goal. We propose a new approach called replication based mitigation (RBM) based on parallel quantum annealing (QA). In RBM, physical qubits representing the same logical qubit are dispersed across different copies of the problem embedded in the hardware. This mitigates hardware biases, is compatible with limited qubit connectivity in current annealers, and is well-suited for currently available noisy intermediate-scale quantum annealers. Our experimental analysis shows that RBM provides solution quality on par with previous methods while being more flexible and compatible with a wider range of hardware connectivity patterns. In comparisons against standard QA without error mitigation on larger problem instances that could not be handled by previous methods, RBM consistently gets better energies and ground state probabilities across parameterized problem sets.

Quantum Science and Technology•Aug 13, 2024•2 citations

Decoherence-free quantum error mitigation by density matrix vectorization

Zhong-Xia Shang, Zihan Chen, Cai-Sheng Cheng

Fighting against noise is crucial for Noisy Intermediate Scale Quantum (NISQ) devices to demonstrate practical quantum applications. In this work, we give a paradigm of quantum error mitigation based on the vectorization of density matrices for variational quantum algorithms. Different from the ideas of existing quantum error mitigation methods that try to distill noiseless information from noisy quantum states, our proposal directly changes the way of encoding information and maps the density matrices of noisy quantum states to noiseless virtual pure states, which is realized by an NISQ-friendly measurement protocol and a classical postprocessing procedure. Our protocol requires no knowledge of the noise model, no ability to tune the noise strength, and no ancilla qubits for complicated controlled unitaries. Under our encoding, NISQ devices are always preparing virtual pure quantum states that are highly desired resources for variational quantum algorithms to have good performance in many tasks. We show how this protocol can be well fitted into variational quantum algorithms. We give several concrete ansatz constructions that are suitable for our proposal and do theoretical analysis on the sampling complexity, the expressibility, and the trainability. We also give a discussion on how this protocol is influenced by large noise and how it can be well combined with other quantum error mitigation protocols. The effectiveness of our proposal is demonstrated by various numerical experiments.

arXiv:2405.07592•Physical Review Research•May 3, 2026•0 citations

Logical shadow tomography: Efficient estimation of error-mitigated observables

Hong-Ye Hu, Ryan LaRose, Yi‐Zhuang You +2 more

We introduce a technique to estimate error-mitigated expectation values on noisy quantum computers. Our technique performs shadow tomography on a logical state to produce a memory-efficient classical reconstruction of the noisy density matrix. Using efficient classical post-processing, one can mitigate errors by projecting a general nonlinear function of the noisy density matrix into the codespace. The subspace expansion and virtual distillation can be viewed as special cases of the new framekwork. We show our method is favorable in the quantum and classical resources overhead. Relative to subspace expansion which requires $O\left(2^{N} \right)$ samples to estimate a logical Pauli observable with $[[N, k]]$ error correction code, our technique requires only $O\left(4^{k} \right)$ samples. Relative to virtual distillation, our technique can compute powers of the density matrix without additional copies of quantum states or quantum memory. We present numerical evidence using logical states encoded with up to sixty physical qubits and show fast convergence to error-free expectation values with only $10^5$ samples under 1% depolarizing noise.

arXiv:2203.07263•Mar 14, 2022•7 citations

Practical quantum error mitigation for analog quantum simulation

Jinzhao Sun, Xiao Yuan, Takahiro Tsunoda +3 more

Analog quantum simulation has been proposed as an efficient approach for probing classically intractable many-body physics. In contrast to digital, gate-based quantum simulation, the analog approach involves configuring the quantum hardware to directly mimic the physics of the target system. This reduces the level of control required and can offer a degree of inherent robustness, but it comes at the cost that known error-handling techniques may be inapplicable. Accumulation of error is thus a major challenge facing medium-scale and large-scale analog quantum simulation with current and near-term noisy quantum (NISQ) devices. Here we propose a hybrid error mitigation scheme that can suppress general local Markovian noise in analog quantum simulators. We employ stochastic error mitigation for physical noise and use Richardson extrapolation to compensate for model estimation error. We numerically test the error mitigation scheme for an Ising-type Hamiltonian with energy relaxation and dephasing noise, and show an improvement of simulation accuracy by two orders. We assess the resource cost of our error mitigation scheme and conclude that analog quantum simulation at scale is feasible with NISQ hardware.

•Jan 14, 2020•1 citation

Replication-Based Quantum Annealing Error Mitigation

Hristo N. Djidjev

Quantum annealers like those from D-Wave Systems implement adiabatic quantum computing to solve optimization problems, but their analog nature and limited control functionalities present challenges to correcting or mitigating errors. As quantum computing advances towards applications, effective error suppression is an important research goal. We propose a new approach called replication based mitigation (RBM) based on parallel quantum annealing. In RBM, physical qubits representing the same logical qubit are dispersed across different copies of the problem embedded in the hardware. This mitigates hardware biases, is compatible with limited qubit connectivity in current annealers, and is suited for available noisy intermediate-scale quantum (NISQ) annealers. Our experimental analysis shows that RBM provides solution quality on par with previous methods while being compatible with a much wider range of hardware connectivity patterns. In comparisons against standard quantum annealing without error mitigation, RBM consistently improves the energies and ground state probabilities across parameterized problem sets.

arXiv:2404.06580•ACM International Conference on Computing Frontiers (CF)•May 7, 2024•1 citation
Quantum annealing (QA) is an efficient method for finding the ground-state energy of the problem Hamiltonian. However, in practical implementation, the system suffers from decoherence. On the other hand, recently, ``Localized virtual purification" (LVP) was proposed to suppress decoherence in the context of noisy intermediate-scale quantum (NISQ) devices. Suppose observables have spatially local support in the lattice. In that case, the requirement for LVP is to calculate the expectation value with a reduced density matrix on a portion of the total system. In this work, we propose a method to mitigate decoherence errors in QA using LVP. The key idea is to use the so-called classical shadow method to construct the reduced density matrix. Thanks to the CS, unlike the previous schemes to mitigate decoherence error for QA, we do not need either two-qubit gates or mid-circuit measurements, which means that our method is hardware-efficient.
Quantum annealing (QA) is one of the methods to prepare a ground state of a problem Hamiltonian. In the absence of noise, QA can accurately estimate the ground-state energy if the adiabatic condition is satisfied. However, in practice, systems are known to suffer from decoherence. Meanwhile, considerable research effort has been devoted to noisy intermediate-scale quantum (NISQ) computation. For practical NISQ computation, many error-mitigation (EM) methods have been proposed to mitigate the effects of noise. This paper proposes a QA strategy combined with an EM method, namely, dual-state purification, to suppress the effects of decoherence. Our protocol consists of four parts: the conventional dynamics, single-qubit projective measurements, the Hamiltonian dynamics corresponding to an inverse map of the first dynamics, and postprocessing of the measurement results. Importantly, our protocol works without two-qubit gates that require pulse operations; hence, it is suitable for devices designed for practical QA. In addition, we present numerical calculations to show that our protocol is more accurate than the conventional QA in estimating the ground-state energy under decoherence.
Abstract Quantum annealers like those manufactured by D-Wave Systems are designed to find high quality solutions to optimization problems that are typically hard for classical computers. They utilize quantum effects like tunneling to evolve toward low-energy states representing solutions to optimization problems. However, their analog nature and limited control functionalities present challenges to correcting or mitigating hardware errors. As quantum computing advances towards applications, effective error suppression is an important research goal. We propose a new approach called replication based mitigation (RBM) based on parallel quantum annealing (QA). In RBM, physical qubits representing the same logical qubit are dispersed across different copies of the problem embedded in the hardware. This mitigates hardware biases, is compatible with limited qubit connectivity in current annealers, and is well-suited for currently available noisy intermediate-scale quantum annealers. Our experimental analysis shows that RBM provides solution quality on par with previous methods while being more flexible and compatible with a wider range of hardware connectivity patterns. In comparisons against standard QA without error mitigation on larger problem instances that could not be handled by previous methods, RBM consistently gets better energies and ground state probabilities across parameterized problem sets.
Fighting against noise is crucial for Noisy Intermediate Scale Quantum (NISQ) devices to demonstrate practical quantum applications. In this work, we give a paradigm of quantum error mitigation based on the vectorization of density matrices for variational quantum algorithms. Different from the ideas of existing quantum error mitigation methods that try to distill noiseless information from noisy quantum states, our proposal directly changes the way of encoding information and maps the density matrices of noisy quantum states to noiseless virtual pure states, which is realized by an NISQ-friendly measurement protocol and a classical postprocessing procedure. Our protocol requires no knowledge of the noise model, no ability to tune the noise strength, and no ancilla qubits for complicated controlled unitaries. Under our encoding, NISQ devices are always preparing virtual pure quantum states that are highly desired resources for variational quantum algorithms to have good performance in many tasks. We show how this protocol can be well fitted into variational quantum algorithms. We give several concrete ansatz constructions that are suitable for our proposal and do theoretical analysis on the sampling complexity, the expressibility, and the trainability. We also give a discussion on how this protocol is influenced by large noise and how it can be well combined with other quantum error mitigation protocols. The effectiveness of our proposal is demonstrated by various numerical experiments.
We introduce a technique to estimate error-mitigated expectation values on noisy quantum computers. Our technique performs shadow tomography on a logical state to produce a memory-efficient classical reconstruction of the noisy density matrix. Using efficient classical post-processing, one can mitigate errors by projecting a general nonlinear function of the noisy density matrix into the codespace. The subspace expansion and virtual distillation can be viewed as special cases of the new framekwork. We show our method is favorable in the quantum and classical resources overhead. Relative to subspace expansion which requires O(2N)O\left(2^{N} \right)O(2N) samples to estimate a logical Pauli observable with [[N,k]][[N, k]][[N,k]] error correction code, our technique requires only O(4k)O\left(4^{k} \right)O(4k) samples. Relative to virtual distillation, our technique can compute powers of the density matrix without additional copies of quantum states or quantum memory. We present numerical evidence using logical states encoded with up to sixty physical qubits and show fast convergence to error-free expectation values with only 10510^5105 samples under 1% depolarizing noise.
Analog quantum simulation has been proposed as an efficient approach for probing classically intractable many-body physics. In contrast to digital, gate-based quantum simulation, the analog approach involves configuring the quantum hardware to directly mimic the physics of the target system. This reduces the level of control required and can offer a degree of inherent robustness, but it comes at the cost that known error-handling techniques may be inapplicable. Accumulation of error is thus a major challenge facing medium-scale and large-scale analog quantum simulation with current and near-term noisy quantum (NISQ) devices. Here we propose a hybrid error mitigation scheme that can suppress general local Markovian noise in analog quantum simulators. We employ stochastic error mitigation for physical noise and use Richardson extrapolation to compensate for model estimation error. We numerically test the error mitigation scheme for an Ising-type Hamiltonian with energy relaxation and dephasing noise, and show an improvement of simulation accuracy by two orders. We assess the resource cost of our error mitigation scheme and conclude that analog quantum simulation at scale is feasible with NISQ hardware.
Quantum annealers like those from D-Wave Systems implement adiabatic quantum computing to solve optimization problems, but their analog nature and limited control functionalities present challenges to correcting or mitigating errors. As quantum computing advances towards applications, effective error suppression is an important research goal. We propose a new approach called replication based mitigation (RBM) based on parallel quantum annealing. In RBM, physical qubits representing the same logical qubit are dispersed across different copies of the problem embedded in the hardware. This mitigates hardware biases, is compatible with limited qubit connectivity in current annealers, and is suited for available noisy intermediate-scale quantum (NISQ) annealers. Our experimental analysis shows that RBM provides solution quality on par with previous methods while being compatible with a much wider range of hardware connectivity patterns. In comparisons against standard quantum annealing without error mitigation, RBM consistently improves the energies and ground state probabilities across parameterized problem sets.